The purpose of this application is to complete the study, "A Method to Predict Fracture in a Rural County" (AR40879). That proposal was funded beginning 1/1/92. It included three successive phases of an epidemiologic study that was planned for five years. The investigators were advised, at the beginning of the funding period, that five year proposals were not being funded at that time and they were encouraged to submit this Competing Continuation request in order to complete the study. The primary aim of the three phase study was to determine whether patellar ultrasound is a predictor of low-trauma fracture. Two of the three phases of the study have been completed according to the time frame outlined in the original proposal. They are described in the Progress Report in this request. The third phase includes a four year prospective study; two of those four years have been completed as of 11/1/94. It took nearly twenty years to determine that bone mass is a predictor of low-trauma fractures because appropriate prospective studies were not implemented to show this relationship. The investigators have initiated a population-based prospective study of people 50 years of age and older to determine the predictive ability of ultrasound. Ultrasonic measurement of bone integrity is an appealing screening device because it is quickly performed, measures without radiation, and is relatively inexpensive to perform. Additionally, the measuring device used by this research center is portable and can be taken to the epidemiologic field. Better estimates of risk can be obtained from population-based studies. Ultrasonic measurement of bone also is interesting because it measures structural properties of bone, primarily stiffness, in addition to bone mass. Therefore it adds a new dimension to the understanding of the bone pathology that results in low-trauma fractures. The investigators have already demonstrated the association between prevalent low-trauma fractures and low bone quality measured by ultrasound. They will not be able to demonstrate the predictive ability of ultrasound unless this study is completed since, according to the original estimates of fracture incidence rates, not enough of the study participants would be expected to have experienced incident fractures in only half of the required study time. The investigators state that this study will determine, in a four year follow-up time, whether ultrasound predicts low-trauma fractures.